|
Activity Number:
|
433
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
ENAR
|
| Abstract - #303552 |
|
Title:
|
A Meta-Analytic Framework for Combining Incomparable Cox Proportional Hazard Models Caused by Omitting Important Covariates
|
|
Author(s):
|
Xing Yuan*+ and Stewart Anderson
|
|
Companies:
|
University of Pittsburgh and University of Pittsburgh
|
|
Address:
|
4733 Centre Ave Apt 2D, Pittsburgh, PA, 15213,
|
|
Keywords:
|
Cox proportional hazard model ; omitted covariates ; meta-analysis ; aggregated patient data ; individual patient data
|
|
Abstract:
|
Estimated treatment effects from Cox models with some covariates omitted will be biased toward zero. This is problematic in meta-analyses to combine parameter estimates from studies with different covariate adjustments. We propose a meta-analytic framework for combining incomparable Cox models for both aggregated patient data (APD) and individual patient data (IPD). For APD, two meta-regression models are proposed to adjust the heterogeneity of treatment effects across studies. Both parametric and nonparametric estimators for the pooled treatment effect and the heterogeneity variance are presented. For IPD, we propose a weighted estimating equation based on frailty models accommodating covariate omission, and it is compared with multiple imputations method. We illustrate the advantages of proposed procedures over existing methodologies by simulation studies and real data analyses.
|
